Incremental value of novel immunonutritional-inflammatory markers over clinical models in predicting sepsis after ureteroscopic lithotripsy.

Abstract

Background: Sepsis is a serious complication after ureteroscopic lithotripsy (URSL), but its accurate pre-operative prediction remains difficult. This study aims to evaluate novel immunonutritional-inflammatory indices, which reflect the "nutrition-immunity-inflammation" axis, for predicting post-URSL sepsis.

Methods: We conducted a retrospective cohort study of 569 patients undergoing URSL. Postoperative sepsis was diagnosed per Sepsis-3 criteria. A base clinical model was built using pre-operative variables from multivariate analysis. We then assessed the incremental predictive value of adding five indices (CALLY, PNI, CAR, NAR, NPAR) by evaluating discrimination (AUC), calibration, reclassification (NRI, IDI), and clinical utility (Decision Curve Analysis). A nomogram and web calculator were developed based on the CALLY-augmented model for clinical use.

Results: Of 569 patients, 41 (7.2%) developed sepsis. The base model (including gender, platelet count, operation time, hydronephrosis, leukocyte esterase, and urine culture) had an AUC of 0.832. The CALLY index provided the most significant improvement, increasing the AUC to 0.918 (P = 0.007). The model with CALLY also showed superior reclassification (NRI = 0.404, IDI = 0.134; both P < 0.001) and significantly better fit (Likelihood Ratio Test P < 0.001). Decision curve analysis confirmed its greater net clinical benefit across practical risk thresholds.

Conclusion: Immunonutritional-inflammatory indices, particularly the CALLY index, significantly improve the prediction of post-URSL sepsis over clinical factors alone. The developed tool enables personalized risk estimation, supporting precision medicine in perioperative care.

Trial registration: Chinese Clinical Trial Registry, ChiCTR2400092429. Registered 15 November 2024. Retrospectively registered.